<div dir="ltr">Hi Hutter,<div> Thank you so much! I followed your suggestions and I finally get some reasonable results though I haven't obtained very good convergence. I try to use the GTH parameters of Ce as initial guess to generate the GTH parameter of Pr. My parameter is:</div><div><div>&POWELL</div><div> ACCURACY 1.e-14</div><div> STEP_SIZE 0.08</div><div> STEP_SIZE_SCALING 0.90</div><div> MAX_INIT 50</div><div> MAX_FUN 100</div><div> WEIGHT_PSIR0 0.0</div><div> TARGET_POT_SEMICORE [eV] 0.003000</div><div> TARGET_POT_VALENCE [eV] 0.000300</div><div> TARGET_POT_VIRTUAL [eV] 0.003000</div><div> SEMICORE_LEVEL [eV] 15.0</div><div> WEIGHT_POT_NODE 10.0</div><div> WEIGHT_POT_SEMICORE 10.0</div><div> WEIGHT_POT_VALENCE 5.0</div><div> WEIGHT_POT_VIRTUAL 2.0</div><div> &END</div><div>&END ATOM</div><div> I give a large STEP_SIZE at first, and then I decrease the STEP_SIZE by order for the following restarts. The optimization results are:</div><div><div> Reference configuration 1 Method number 1</div><div> L N Occupation Eigenvalue [eV] dE [eV] dCharge</div><div> 0 1 2.00 -38.8397939312 SC -0.009395[ 1] -0.000503[ X]</div><div> 0 2 2.00 -3.3860472487 VA 0.010363[87] -0.004646[ 0]</div><div> 0 3 0.00 1.6037528306 U1 0.013647[ 0] -0.001261[ X]</div><div> 0 4 0.00 8.1521110396 U2 0.035468[ 0] -0.001273[ 0]</div><div> 1 1 6.00 -21.3735479056 SC 0.024748[ 8] -0.002697[ X]</div><div> 1 2 0.00 -1.0360212280 U1 0.029883[ 2] -0.009428[ 0]</div><div> 1 3 0.00 3.6087950888 U2 0.064960[ 0] -0.005279[ 0]</div><div> 2 1 0.00 -1.9516163818 U1 0.002300[ X] -0.014541[ 0]</div><div> 2 2 0.00 2.1059505831 U2 0.025993[ X] 0.000310[ 0]</div><div> 3 1 3.00 -1.2120770213 VA 0.000989[ 0] -0.005278[ 0]</div><div> 3 2 0.00 2.9876786134 U1 -0.004826[ 0] 0.001172[ X]</div><div> 3 3 0.00 8.1202646230 U2 0.034044[ 0] -0.006754[ 0]</div><div> s-states N= 1 Wavefunction at r=0: 0.002169[ 0]</div><div> s-states N= 2 Wavefunction at r=0: 0.018634[ 0]</div><div> s-states N= 3 Wavefunction at r=0: 0.016484[ 0]</div><div> s-states N= 4 Wavefunction at r=0: -0.030807[ 0]</div><div><br></div><div> Number of target values reached: 6 of 20</div></div><div> It still needs several steps restart. I have a question about the "optimal parameter". what is the standard for us to choose the best parameter? I just try to find the best parameter which can make the objective function lowest for each optimization.</div><div> With my best regards,</div><div> Zhi</div><br>在 2018年12月21日星期五 UTC-8上午1:12:02，jgh写道：<blockquote class="gmail_quote" style="margin: 0;margin-left: 0.8ex;border-left: 1px #ccc solid;padding-left: 1ex;">Hi <br> <br>as said before, there is no unique solution to the problem. <br>I gave you the parameters I use, but you can change them at your <br>will. Use what works best for you. <br>Yes, I also use frequent restarts with adjustments of parameters. <br> <br>regards <br> <br>Juerg Hutter <br>------------------------------<wbr>------------------------------<wbr>-- <br>Juerg Hutter Phone : ++41 44 635 4491 <br>Institut für Chemie C FAX : ++41 44 635 6838 <br>Universität Zürich E-mail: <a href="javascript:" target="_blank" gdf-obfuscated-mailto="KJFojUpNCAAJ" rel="nofollow" onmousedown="this.href='javascript:';return true;" onclick="this.href='javascript:';return true;">hut...@chem.uzh.ch</a> <br>Winterthurerstrasse 190 <br>CH-8057 Zürich, Switzerland <br>------------------------------<wbr>------------------------------<wbr>--- <br> <br>-----<a href="javascript:" target="_blank" gdf-obfuscated-mailto="KJFojUpNCAAJ" rel="nofollow" onmousedown="this.href='javascript:';return true;" onclick="this.href='javascript:';return true;">cp...@googlegroups.com</a> wrote: ----- <br>To: "cp2k" <<a href="javascript:" target="_blank" gdf-obfuscated-mailto="KJFojUpNCAAJ" rel="nofollow" onmousedown="this.href='javascript:';return true;" onclick="this.href='javascript:';return true;">cp...@googlegroups.com</a>> <br>From: <a href="javascript:" target="_blank" gdf-obfuscated-mailto="KJFojUpNCAAJ" rel="nofollow" onmousedown="this.href='javascript:';return true;" onclick="this.href='javascript:';return true;">ning...@gmail.com</a> <br>Sent by: <a href="javascript:" target="_blank" gdf-obfuscated-mailto="KJFojUpNCAAJ" rel="nofollow" onmousedown="this.href='javascript:';return true;" onclick="this.href='javascript:';return true;">cp...@googlegroups.com</a> <br>Date: 12/21/2018 07:09AM <br>Subject: Re: [CP2K:11071] Generate the GTH pseudopotential by ATOM <br> <br>Hi Hutter, <br> Thank you! I try to change the STEP_SIZE and MAX_FUN. I use the optimal parameter, and it can work. <br> There is another problem. We will choose a series parameters as initial guess for pseudopotential optimization. Then we use the so-called optimal STEP_SIZE and MAX_FUN. But we always need to restart the optimization using the pseudopotential parameter of previous step as initial guess. Do we still use the same STEP_SIZE and MAX_FUN? I just use the same value. But I think we need to use adaptive STEP_SIZE. I know it is very hard to change the STEP_SIZE for each restart. <br> By the way, I think we should also test the weight for each item in objective function. <br> With my best regards, <br> Zhi <br> <br>在 2018年12月19日星期三 UTC-8上午12:45:36，jgh写道：Hi <br> <br>as I said in the last comment - it is an art - to choose the <br>best parameter. You need some experience to get "optimal" <br>performance/convergence. <br>I usually play with these parameters <br> <br> STEP_SIZE 0.08 <br> MAX_INIT 50 <br> MAX_FUN 250 <br> STEP_SIZE_SCALING 0.90 <br> <br>Start with large STEP_SIZE (0.1) and small MAX_FUN (50-100) <br>and end with small STEP_SIZE (0.00001) and large MAX_FUN(5000). <br> <br>regards <br> <br>Juerg Hutter <br>------------------------------<wbr>------------------------------<wbr>-- <br>Juerg Hutter Phone : ++41 44 635 4491 <br>Institut für Chemie C FAX : ++41 44 635 6838 <br>Universität Zürich E-mail: <a>hut...@chem.uzh.ch</a> <br>Winterthurerstrasse 190 <br>CH-8057 Zürich, Switzerland <br>------------------------------<wbr>------------------------------<wbr>--- <br> <br>-----<a>cp...@googlegroups.com</a> wrote: ----- <br>To: "cp2k" <<a>cp...@googlegroups.com</a>> <br>From: <a>ning...@gmail.com</a> <br>Sent by: <a>cp...@googlegroups.com</a> <br>Date: 12/18/2018 08:15PM <br>Subject: Re: [CP2K:11058] Generate the GTH pseudopotential by ATOM <br> <br>Hi Hutter, <br> I find I need to restart the optimization for many times. It always needs 5000 restarting calculations using previous step GTH parameters. This process is so slow. How can I speed up this process? <br> With my best regards, <br> Zhi <br> <br>在 2018年12月12日星期三 UTC-8下午11:56:59，jgh写道：Hi <br> <br>for many atoms you have to restart the optimization frequently <br>until you reach sufficient convergence. <br>For the restart you copy the GTH-PARAMETER to your input. <br>Optimization is a kind of art. Choosing the parameter in the <br>POWELL section is important for good performance. Still, it takes <br>me often several 10'000 steps to reach reasonable results. <br> <br>regards <br> <br>Juerg Hutter <br>------------------------------<wbr>------------------------------<wbr>-- <br>Juerg Hutter Phone : ++41 44 635 4491 <br>Institut für Chemie C FAX : ++41 44 635 6838 <br>Universität Zürich E-mail: <a>hut...@chem.uzh.ch</a> <br>Winterthurerstrasse 190 <br>CH-8057 Zürich, Switzerland <br>------------------------------<wbr>------------------------------<wbr>--- <br> <br>-----<a>cp...@googlegroups.com</a> wrote: ----- <br>To: "cp2k" <<a>cp...@googlegroups.com</a>> <br>From: <a>ning...@gmail.com</a> <br>Sent by: <a>cp...@googlegroups.com</a> <br>Date: 12/12/2018 06:39PM <br>Subject: Re: [CP2K:11046] Generate the GTH pseudopotential by ATOM <br> <br>Hi Hutter, <br> Thank you! I use your input file to generate the pseudopotential of Ti and I can reproduce your result. But I still have a question, the initial guess is nearly same with the final optimized pseudopotential. How can we get the better initial guess? These parameters include the R^cut and coefficients in non-local part and R^cut for each channel and h^ij coefficients in the local part. <br> With my best regards, <br> Zhi <br> <br>在 2018年12月12日星期三 UTC-8上午12:08:27，jgh写道：Hi <br> <br>the optimization of GTH pseudos is not unique. Depending on your <br>object function, starting point and convergence you can get rather <br>different results. The performance of all these different pseudos <br>should be rather similar though. <br>There are also elements that can be tricky to get and many repeated <br>optimizations and some 'magic' is needed to get good results. <br> <br>This is a generic input I am currently using to optimize potentials: <br> <br>&GLOBAL <br> PROGRAM_NAME ATOM <br>&END GLOBAL <br>&ATOM <br> ELEMENT Ti <br> <br> RUN_TYPE PSEUDOPOTENTIAL_OPTIMIZATION <br> <br> ELECTRON_CONFIGURATION [Ne] 3s2 3p6 4s2 3d2 <br> CORE [Ne] <br> MAX_ANGULAR_MOMENTUM 3 <br> <br> &METHOD <br> METHOD_TYPE KOHN-SHAM <br> RELATIVISTIC DKH(3) <br> &XC <br> &XC_FUNCTIONAL <br> &LIBXC <br> FUNCTIONAL MGGA_X_SCAN <br> &END LIBXC <br> &LIBXC <br> FUNCTIONAL MGGA_C_SCAN <br> &END LIBXC <br> &END XC_FUNCTIONAL <br> &END XC <br> &END METHOD <br> &OPTIMIZATION <br> EPS_SCF 1.e-7 <br> &END <br> &AE_BASIS <br> BASIS_TYPE GEOMETRICAL_GTO <br> &END AE_BASIS <br> &PP_BASIS <br> BASIS_TYPE GEOMETRICAL_GTO <br> &END PP_BASIS <br> &POTENTIAL <br> PSEUDO_TYPE GTH <br> &GTH_POTENTIAL <br> 4 6 2 0 <br> 0.38300965389957 2 8.69526962380773 -0.69130398136282 <br> 3 <br> 0.32535306838298 2 2.49370385691556 3.69297179681161 <br> -4.49419494638446 <br> 0.25460783706014 2 -4.62952832752021 8.87087769105041 <br> -10.49552627066998 <br> 0.24470423408872 1 -9.41064468800323 <br> &END <br> CONFINEMENT_TYPE BARRIER <br> CONFINEMENT 200. 4.0 12.0 <br> &END POTENTIAL <br> <br> &POWELL <br> ACCURACY 1.e-14 <br> STEP_SIZE 0.08 <br> MAX_INIT 50 <br> MAX_FUN 250 <br> STEP_SIZE_SCALING 0.90 <br> WEIGHT_PSIR0 0.0 <br> TARGET_POT_SEMICORE [eV] 0.003000 <br> TARGET_POT_VALENCE [eV] 0.000300 <br> TARGET_POT_VIRTUAL [eV] 0.003000 <br> WEIGHT_POT_NODE 10.0 <br> WEIGHT_POT_SEMICORE 2.0 <br> WEIGHT_POT_VALENCE 5.0 <br> WEIGHT_POT_VIRTUAL 1.0 <br> &END <br>&END ATOM <br> <br>You can find my recent optimized potentials for PBE, PBE0, SCAN, HF <br>at my github repo (absolutely no warrenty!) <br> <br><a href="https://github.com/juerghutter/GTH" target="_blank" rel="nofollow" onmousedown="this.href='https://www.google.com/url?q\x3dhttps%3A%2F%2Fgithub.com%2Fjuerghutter%2FGTH\x26sa\x3dD\x26sntz\x3d1\x26usg\x3dAFQjCNFUAfC8QPoR9bVb_LWRHynmLIMC5g';return true;" onclick="this.href='https://www.google.com/url?q\x3dhttps%3A%2F%2Fgithub.com%2Fjuerghutter%2FGTH\x26sa\x3dD\x26sntz\x3d1\x26usg\x3dAFQjCNFUAfC8QPoR9bVb_LWRHynmLIMC5g';return true;">https://github.com/<wbr>juerghutter/GTH</a> <br> <br>regards <br> <br>Juerg Hutter <br>------------------------------<wbr>------------------------------<wbr>-- <br>Juerg Hutter Phone : ++41 44 635 4491 <br>Institut für Chemie C FAX : ++41 44 635 6838 <br>Universität Zürich E-mail: <a>hut...@chem.uzh.ch</a> <br>Winterthurerstrasse 190 <br>CH-8057 Zürich, Switzerland <br>------------------------------<wbr>------------------------------<wbr>--- <br> <br>-----<a>cp...@googlegroups.com</a> wrote: ----- <br>To: "cp2k" <<a>cp...@googlegroups.com</a>> <br>From: <a>ning...@gmail.com</a> <br>Sent by: <a>cp...@googlegroups.com</a> <br>Date: 12/12/2018 02:19AM <br>Subject: [CP2K:11041] Generate the GTH pseudopotential by ATOM <br> <br>Hello, <br> I need to generate the GTH pseudopotential and hope to use ATOM code in CP2K. But I find it is a little hard for me. The only thing I can do is just changing the initial guess of pseudopotential parameters. But there are so many parameters in GTH pseudopotential. The optimized pseudopotential changes slightly compared with initial guess. Can anyone tell me how to generate the GTH pseudopotential correctly by ATOM? I will be very appreciated! <br> Best regards, <br> Zhi <br> -- <br> You received this message because you are subscribed to the Google Groups "cp2k" group. <br> To unsubscribe from this group and stop receiving emails from it, send an email to <a>cp2k+...@googlegroups.com</a>. <br> To post to this group, send email to <a>cp...@googlegroups.com</a>. <br> Visit this group at <a href="https://groups.google.com/group/cp2k" target="_blank" rel="nofollow" onmousedown="this.href='https://groups.google.com/group/cp2k';return true;" onclick="this.href='https://groups.google.com/group/cp2k';return true;">https://groups.google.com/<wbr>group/cp2k</a>. <br> For more options, visit <a href="https://groups.google.com/d/optout" target="_blank" rel="nofollow" onmousedown="this.href='https://groups.google.com/d/optout';return true;" onclick="this.href='https://groups.google.com/d/optout';return true;">https://groups.google.com/d/<wbr>optout</a>. <br> <br> <br> -- <br> You received this message because you are subscribed to the Google Groups "cp2k" group. <br> To unsubscribe from this group and stop receiving emails from it, send an email to <a>cp2k+...@googlegroups.com</a>. <br> To post to this group, send email to <a>cp...@googlegroups.com</a>. <br> Visit this group at <a href="https://groups.google.com/group/cp2k" target="_blank" rel="nofollow" onmousedown="this.href='https://groups.google.com/group/cp2k';return true;" onclick="this.href='https://groups.google.com/group/cp2k';return true;">https://groups.google.com/<wbr>group/cp2k</a>. <br> For more options, visit <a href="https://groups.google.com/d/optout" target="_blank" rel="nofollow" onmousedown="this.href='https://groups.google.com/d/optout';return true;" onclick="this.href='https://groups.google.com/d/optout';return true;">https://groups.google.com/d/<wbr>optout</a>. <br> <br> <br> -- <br> You received this message because you are subscribed to the Google Groups "cp2k" group. <br> To unsubscribe from this group and stop receiving emails from it, send an email to <a>cp2k+...@googlegroups.com</a>. <br> To post to this group, send email to <a>cp...@googlegroups.com</a>. <br> Visit this group at <a href="https://groups.google.com/group/cp2k" target="_blank" rel="nofollow" onmousedown="this.href='https://groups.google.com/group/cp2k';return true;" onclick="this.href='https://groups.google.com/group/cp2k';return true;">https://groups.google.com/<wbr>group/cp2k</a>. <br> For more options, visit <a href="https://groups.google.com/d/optout" target="_blank" rel="nofollow" onmousedown="this.href='https://groups.google.com/d/optout';return true;" onclick="this.href='https://groups.google.com/d/optout';return true;">https://groups.google.com/d/<wbr>optout</a>. <br> <br> <br> -- <br> You received this message because you are subscribed to the Google Groups "cp2k" group. <br> To unsubscribe from this group and stop receiving emails from it, send an email to <a href="javascript:" target="_blank" gdf-obfuscated-mailto="KJFojUpNCAAJ" rel="nofollow" onmousedown="this.href='javascript:';return true;" onclick="this.href='javascript:';return true;">cp2k+uns...@googlegroups.<wbr>com</a>. <br> To post to this group, send email to <a href="javascript:" target="_blank" gdf-obfuscated-mailto="KJFojUpNCAAJ" rel="nofollow" onmousedown="this.href='javascript:';return true;" onclick="this.href='javascript:';return true;">cp...@googlegroups.com</a>. <br> Visit this group at <a href="https://groups.google.com/group/cp2k" target="_blank" rel="nofollow" onmousedown="this.href='https://groups.google.com/group/cp2k';return true;" onclick="this.href='https://groups.google.com/group/cp2k';return true;">https://groups.google.com/<wbr>group/cp2k</a>. <br> For more options, visit <a href="https://groups.google.com/d/optout" target="_blank" rel="nofollow" onmousedown="this.href='https://groups.google.com/d/optout';return true;" onclick="this.href='https://groups.google.com/d/optout';return true;">https://groups.google.com/d/<wbr>optout</a>. <br> <br></blockquote></div></div>